This article is more than 1 year old

What is HPC actually good for? Just you wait and see

Technology, if not money, does trickle down to your children

Wider access, better ideas

Take or leave Simon's cornucopian economics as you wish, but this idea seems to me to be obvious. There's seven billion of us floating around the planet today and it would be ridiculous to claim that all are provided with the tools that they could use.

It just isn't true that every megabrain on the planet is already using those cutting-edge tools: just as it wasn't true five centuries ago that everyone capable of reading and writing could (that's not even true today).

As these tools become ever cheaper then an ever-larger portion of the global population will have access to them. And as that happens then there's going to be people who don't currently have access to them who come up with bright ideas that benefit us.

And that, in perhaps my slightly warped view, is what the value of this HPC work is. Sure, it's nice to know how proteins are folded. But the techniques and the technologies we develop to produce those petaflops will, soon enough, be on a $25 chip that is incorporated into standard computing devices.

And that's when things will get really interesting, when billions of people (yes, including billions minus a handful who have no idea at all) have that technology then that handful will, undoubtedly, astonish us all with the fruits of their ingenuity.

That development frontier might best be viewed as being useful not for the problems they're working on at that frontier but for the later trickle down effects of the technologies they develop at that frontier.

If you prefer, HPC is really R&D into the techniques of computing and as such is a bargain for the rest of us. Or perhaps, if we're honest about it, for our children. ®

More about

TIP US OFF

Send us news


Other stories you might like